repeated measures help

wsj

New Member
#1
Hello,

I have been trying to figure this out on my own for a couple of days, but have hit the road block. My problem relates to a data set of repeated measures that I am trying to analyze using SPSS ver. 10.

I have performed an experiment looking at intestinal permeability at 15 serial time points. My subjects belong to one of 6 groups that correspond to a type of treatment. There was no crossover between the treatment groups. It would seem that I need a repeated measures GLM with 1 within-subjects factor (time point) that has 15 levels, and 1 between-subjects factor (treatment arm) that has 6 levels. The data is entered into SPSS using 15 columns that correspond to the different time points. I have one final column that contains the grouping variable (1 through 6). The setup seems correct as when SPSS computes the descriptive statistics, it knows how to break my data down and calculate means, stdev, etc., for the individual time points within the groups. I am also seeing the output I would expect, showing that time and time*group interactions are significant in my data. In the post hoc analysis, I can see that overall, group 1 is different from group 6 and so forth. But, I need to be able to go one step further.

Graphically, my data is very similar at the initial time points, but then the groups become hugely different about halfway through the experiment. What I would like to analyze are the individual time points so that I can tell when the groups become significantly different, not just that they are different overall. I am contemplating breaking my data down into 15 subsets, then running multiple pairwise comparisons between my 6 groups using ANOVA, for each time point. However, I suspect this is incorrect as it will not account for the fact that the data comes from subjects that had repeated measures. Any suggestions on how I can manage this analysis?

Thanks